Multi-modal computer-aided diagnosis systems and methods for prostate cancer

Abstract

Methods and apparatus for computer-aided prostate condition diagnosis are disclosed. An example computer-aided prostate condition diagnosis apparatus includes a memory to store instructions and a processor. The example processor is to execute the instructions to implement at least a prostate assessor, a lesion assessor, and an outcome generator. The example prostate assessor is to evaluate a volume and density of a prostate gland in an image of a patient to determine a prostate-specific antigen level for the prostate gland. The example lesion assessor is to analyze a lesion on the prostate gland in the image. The example outcome generator is to generate an assessment of prostate gland health based on the prostate-specific antigen level and the analysis of the lesion.

What is claimed is: 
     
       1. A computer-aided prostate condition diagnosis apparatus comprising:
 a memory to store instructions; and 
 a processor to execute the instructions to implement at least:
 a prostate assessor to evaluate a volume and density of a prostate gland in an image of a patient using distances deposited on the image to determine a prostate specific antigen level for the prostate gland; 
 a lesion assessor to segment a lesion on the prostate gland in the image by depositing an ellipse on the lesion in the image, wherein the lesion assessor is to register the image including the deposited ellipse and a prostate sector map, wherein the lesion assessor is to automatically select one or more sectors corresponding to the ellipse on the segmented lesion, the one or more automatically selected sectors corresponding to a position of the ellipse in the registered prostate sector map, the prostate sector map comprising at least one of an axial plane, a sagittal plane, or a coronal plane, and wherein the lesion assessor is to determine a score representing an analysis of the segmented lesion; and 
 an outcome generator to generate an assessment of prostate gland health based on the prostate-specific antigen level, the score representing the analysis of the segmented lesion, the prostate specific antigen level, the automatically selected one or more sectors, and the analysis of the segmented lesion applied to a machine learning model to generate a value for the assessment of prostate gland health. 
 
 
     
     
       2. The apparatus of  claim 1 , wherein the lesion assessor is to analyze the lesion using a plurality of scores to form a global score to classify the lesion. 
     
     
       3. The apparatus of  claim 1 , wherein the prostate assessor is to evaluate the volume and density of the prostate gland by segmenting and modeling the image using a deep learning network model. 
     
     
       4. The apparatus of  claim 1 , wherein the lesion assessor is to analyze the lesion by segmenting and modeling the image using a deep learning network model. 
     
     
       5. The apparatus of  claim 1 , wherein the lesion assessor is to overlay the ellipse on the lesion, the overlay of the ellipse to trigger the automatic selection of the one or more sectors underneath the ellipse in the image. 
     
     
       6. The apparatus of  claim 1 , wherein the image includes a three-dimensional volume. 
     
     
       7. The apparatus of  claim 1 , wherein the prostate assessor is to model the prostate gland with a digital twin. 
     
     
       8. A non-transitory computer-readable storage medium including instructions which, when executed, cause at least one processor to at least:
 evaluate a volume and density of a prostate gland in an image of a patient using distances deposited on the image to determine a prostate-specific antigen level for the prostate gland; 
 segment a lesion on the prostate gland in the image by depositing an ellipse on the lesion in the image, 
 register the image including the deposited ellipse and a prostate sector map, 
 automatically select one or more sectors corresponding to the ellipse on the segmented lesion, the one or more automatically selected sectors corresponding to a position of the ellipse in the registered prostate sector map, the prostate sector map comprising at least one of an axial plane, a sagittal plane, or a coronal plane; 
 determine a score representing an analysis of the segmented lesion; and 
 generate an assessment of prostate gland health based on the prostate-specific antigen level, the score representing the analysis of the segmented lesion, the prostate-specific antigen level, the automatically selected one or more sectors, and the analysis of the segmented lesion applied to a machine learning model to generate a value for the assessment of prostate gland health. 
 
     
     
       9. The computer-readable storage medium of  claim 8 , wherein the instructions, when executed, cause the at least one processor to analyze the lesion using a plurality of scores to form a global score to classify the lesion. 
     
     
       10. The computer-readable storage medium of  claim 8 , wherein the instructions, when executed, cause the at least one processor to evaluate the volume and density of the prostate gland by segmenting and modeling the image using a deep learning network model. 
     
     
       11. The computer-readable storage medium of  claim 8 , wherein the instructions, when executed, cause the at least one processor to analyze the lesion by segmenting and modeling the image using a deep learning network model. 
     
     
       12. The computer-readable storage medium of  claim 8 , wherein the instructions, when executed, cause the at least one processor to overlay the ellipse on the lesion to trigger the automatic selection of the one or more sectors underneath the ellipse in the image. 
     
     
       13. The computer-readable storage medium of  claim 8 , wherein the image includes a three-dimensional volume. 
     
     
       14. The computer-readable storage medium of  claim 8 , wherein the instructions, when executed, cause the at least one processor to model the prostate gland with a digital twin. 
     
     
       15. A method for computer-aided prostate condition diagnosis, the method comprising:
 evaluating, with at least one processor, a volume and density of a prostate gland in an image of a patient using distances deposited on the image to determine a prostate specific antigen level for the prostate gland; 
 segmenting, with the at least one processor, a lesion on the prostate gland in the image by depositing an ellipse on the lesion in the image; 
 registering the image including the deposited ellipse and a prostate sector map, 
 automatically selecting one or more sectors corresponding to the ellipse on the segmented lesion, the one or more sectors corresponding to a position of the ellipse in the registered prostate sector map, the prostate sector map comprising at least one of an axial plane, a sagittal plane, or a coronal plane; 
 determining a score representing an analysis of the segmented lesion; and 
 generating, with the at least one processor, an assessment of prostate gland health based on the prostate-specific antigen level, the score representing the analysis of the segmented lesion, the prostate specific antigen level, the automatically selected one or more sectors, and the analysis of the segmented lesion applied to a machine learning model to generate a value for the assessment of prostate gland health. 
 
     
     
       16. The method of  claim 15 , wherein analyzing the lesion includes analyzing the legion using a plurality of scores to form a global score to classify the lesion. 
     
     
       17. The method of  claim 15 , wherein evaluating the volume and density of the prostate gland includes evaluating the volume and density of the prostate gland by segmenting and modeling the image using a deep learning network model. 
     
     
       18. The method of  claim 15 , wherein analyzing the lesion includes analyzing the lesion by segmenting and modeling the image using a deep learning network model. 
     
     
       19. The method of  claim 15 , further including overlaying the ellipse on the lesion to trigger the automatic selection of the one or more sectors underneath the ellipse in the image. 
     
     
       20. The method of  claim 15 , wherein the image includes a three-dimensional volume.